Cloud and Mobile App Performance
summary
A demo cloud+mobile app uploads a phone photo to a browser via a Google Cloud backend and instruments a waterfall trace to locate real bottlenecks. Eight concrete fixes are applied: remove redundant database calls, defer cleanup work, batch and parallelize operations, stream data, decompose tasks, relax strict requirements, use low‑level I/O, and tune the UX; measured improvements include resizing >5 MB images to ~200 KB (~15x), creating a 7 KB preview, avoiding a network bug that caused ~6+ seconds/MB from 32 KB chunking, and reducing final handoff latency to about 150 ms.
AI-written summary · about
chapters (7)
- 0:00Start
- 7:31Hunt unnecessary operations
- 10:15Batching
- 11:26Parallelize operations
- 16:24Dig deeper
- 17:48Revisit requirements
- 19:27Mechanical sympathy
related talks
37mIt's just too Slow: Testing Mobile application performance
37mMobile Web Performance
33mOptimization - Making your app faster
08mThe Performance Alphabet
39mPushing the Limits of Mobile Performance
41mModern App Performance
29mClient Says Your Awesome Web App Works Slowly. Now What?
41mBuild apps for Billions